Quantum Electrodynamics is rotten at the core

Quantum electrodynamics is considered the most accurate theory in the history of science. This precision is all based on a single experimental value - the anomalous magnetic moment of the electron called the g-factor. In this episode, I want to examine a paper by Oliver Consa who examines the very suspicious coincidences, errors, mathematical inconsistencies and renormalisation infinities which have been swept under the rug.

Please consider supporting this channel through:

or CRYPTO Donations:
Bitcoin: bc1q5cctzkc9tt6hmqueddfk5dlvcpr6y45gx7td04
Ethereum: 0x2df869b96d4b42c461635B2955fAF72C79eA445D
Dogcoin: DRUEVXavwhbavuhgYJV2AXo8N6tC2zB5za

Other Videos Relating to Quantum Mechanics:

Follow me on:

References:

00:00 Introduction
01:54 Manhattan Project
03:46 Dirac's equation
04:38 Quantum Field Theory and Ignoring Infinities
05:57 Shelter Island Conference
07:43 Bethe's Lamb Shift
08:19 Schwinger factor
09:50 2nd Conference
12:08 Dyson's Unification
13:55 3rd Conference
15:40 Dyson points out divergence after normalisation
16:31 Doctoring theoretical value to match experiment
18:04 Coefficient rabbit hole
24:12 Muon's g-factor problem
25:14 Fudging the electron g-factor
26:24 Final remarks

[michaelconvery]

QED isn't perfect, agreed. But I think there's an unjustifiable gap in the chronology of your presentation, and this omission severely damages the claims of your argument. You skip the crucial developments of the 1970s and 80s. First, it was shown by Glashow, Salam, and Weinberg that QED is an effective field theory, an approximation of the electroweak theory that's only valid in a limited range of energy scales. Second, and perhaps more importantly, there's the more general work of Ken Wilson who discovered the physical rational behind renormalization. He showed that renormalization, rather than being a method of hand-waving infinities under the rug, was a completely justifiable mathematical technique which reflected how the phenomena of field theories are organized by scale. Renormalization doesn't just work for QED, it works for higher energy theories like QCD and the electroweak theory, but it also works for the lower energy systems of condensed matter. (Try telling a condensed matter physicist to stop using renormalization, and I guarantee it won't end well)

My point is this:
Because your presentation doesn't take into account the more modern view of renormalization and of QED as an effective field theory, it comes across as an insufficient and misinformed suspicion towards a list of coefficients.

Yes, the numbers disagree. But only at the 10th decimal place ! You really need to emphasize how tiny that is, and then dive into all the many possible complicated factors that could account for those tiny discrepancies, before you advocate we ditch QED.

EDIT: For those interested, the resources that helped me with some of these ideas are Polchinski's TASI notes "Effective Field Theory and the Fermi Surface", David Tong's notes "Statistical Field Theory", and the 8th lecture from Tobias Osborne's advanced QFT course uploaded to Youtube entitled "Watch this first! Advanced quantum field theory, Lecture 8". I hope those help

[Bankoru]

This is sweeping a bit under the rug why renormalization is "wrong", or rather, why you shouldn't just discard all current theories based on a lack of knowledge. It is not that it is a lie, but rather that it needs more insight and restructuring just as any other previous model in science. The fact that renormalization results in agreements with experiments is simply a hint that more insight is needed. My PhD thesis was somewhat related to this and I feel it's important to understand that nothing we have currently is set in stone.

[herzogsbuick]

"Others gave talks to show how to do the calculations, while Schwinger gave talks to show that only he could do it." I laughed out loud at that, well put Oppenheimer

[tomaschlouba]

I dont want to sound mean, I like that the author did a lot of research to do this video. However large parts are quite misinformed because it is looking on QED from 70 year old viewpoint and some points are mixing different unrelated things together. Most, if not all, these "issues" and discrepancies in theory are now well understood and resolved. QED is not incomplete or ill defined and Schwinger, Tomogana and Feynmans aproaches are all valid. And most importantly: Physics is not based on 50 year old quotes, many of which are out of date or out of context.

What needs to be understood is that f-diagrams are not magic. They are simply graphical representation of series expansion of evolution operator. This is mathematicaly riggorous. What is also well understood now is that these series will diverge due to mathematical properties of Taylor series, essentialy this expansion neglects quantum tunneling. We know for example QED will give you better and better results until terms of power of roughly 137, then terms start to diverge. Exact solution to QED equation of course does not suffer from this problem but is much harder/impossible to obtain.

It has also been said "coupling constant for fermions is greater than one, therefore Dyson series diverges". This is simply not true. Coupling constant of what? Which field? Coupling constant for QED is always 1/137 (+running of course). It is low energy QCD and other high energy QFT where couplings are strong. Then you cannot do Feynman diagrams and have to rely on different, usually numerical aproaches.

Later author introduces muon g-factor. The he claims on the screen that there are two contributions to g value, that is g from QED and g from nonQED. In that g nonQED he introduces g from electroweak interaction. Electroweak interaction is just generalizastion of QED, it contains both QED and weak interactions so this is either misspeak or misunderstanding. What is also misunderstanding is that he then compares muon g factor and electron g factor and claims they should be the same? No, they shouldnt. Actualy discrepancy of experiment and theory for muon g factor is the biggest window into physics beyond the standard model we currently have. So its not like anytime there is discrepancy we just "modify" theory results to make it fit. Quite opposite, we celebrate anytime experiment does not match theory because that is the possibility to discover something new. Today the g-2 precision is up to 10^-10 or so for almost 20 years. There is no discrepancy. Experiment (2006) can verify this up to fifth order, higher orders are too small to be measured.

Also, renormalization and regularization are perfectly fine. You expect infinities in theory which has interaction of 1/r, that is not surprising. I could go deeper here but that would be too long. Hope this helps.

[Seagaltalk]

Yep let's listen to one crank over the actual consensus and experimental evidence

[truebaran]

I don't like the way you presented this issues-like if it were some sort of a conspiracy theory. You didn;t mention Wilson's work which gave conceptual reason for renormalization-you quoted Dirac, Feynmann and so on, which were brilliant but some time has passed since then and now we understand things better than in the fifties. Saying that renormalization is just ignoring the infinities is definitively not saying the whole story: there is a plenty of work about nonperturbative renormalization-for example the recent fields medal was awarded to Hugo Duminil Copin for showing (among other things) triviality of phi^4 model. There are lectures on IHES youtube channel about this. Renormalization techniques are used in the theory of the so called regularity structures which has plenty of application in stochastic partial differential equations (see the work of Martin Hairer and his collaborators). There is a very rich structure behind the renormalization techniques, the so called Hopf algebra of rooted trees and more general of Feynman diagrams-see the work of Connes and Kreimer. Even more, there is some universal group behind the renormalization which was conjectured by Cartier and found in the work of Connes and Marcolli (the so called Cosmic Galois Group). This clearly shows that we understand SOMETHING about renormalization. Do we understand all that can be understood? Of course not, QFT is very tricky buisness but nobody is pretending that there are no problems: it is a shame that you are presenting these issues like if everybody was hiding something. I read just a little bit about these things in Folland's excellent book, ,Quantum Field Theory: A Tourist Guide for Mathematicians'' and nobody is pretending that QFT is well defined mathematically. Everything is stated clearly: that first you need to get rid of infinities using renormalization but even after that the series which you obtain are believed to be divergent (for some of them this is actually proved). It is not the case that these issues are not recognized: in fact, one of the Millenium Prize Problems deals with constructing QFT rigorously.
What I liked in your video is that it serves as a nice summary of the complicated history and highlights how the brighest people are confused about there extremely complicated problems-sine the nature is far more complex that we could imagine

[simongross]

This is very interesting. In another field of study, this would be called "curve fitting". I've always admired Feynman, but after seeing this, I have the impression that Dyson is the most honest of all these guys.

[robind]

This whole thing feels like the professional equivalent of a high school lab class, where students fudged with values to get outcomes like those expected. very weird to see

[arthurgonyeajr]

For a theory allegedly built on a lie it’s funny how we have developed working quantum computers. Or we could just wipe the slate clean go back to the cozy shelter of Newtonian classical physics, and forget all this ever happened.

[koenraad]

It is no wonder Dyson did not get the Nobel price: he was honest. Thank you Dr Consa for the great detective work on QED. Thank you for making this video. The theory is more or less a 'prince of darkness'.

[jchanning]

I have a PhD in Quantum Physics (University of Manchester, 1998) and graduated alongside Prof Brian Cox. Looking back, it staggers me, that 25 years later, people are still "researching" the same little niche area I worked on in the mid-1990s. Most of this research has achieved nothing meaningful, yet people have dedicated their entire lives to it. I am so glad I left this behind and focused on something more practical and real.

[eveningstarnm]

"Quantum mechanics is the idea of the absurd"? I can't think of a reason to listen beyond that first sentence. This channel is for people who don't want to know mathematics and are willing to sound stupid to people who do. PLEASE stop thinking that your ignorance is intelligence. It isn't.

[GamingDemiurge]

When they showed me renormalization techniques I remember calling bs on them. The more I studied the more bs I saw. It is amazing to me how other physicist learn to fool themselves into thinking there is nothing wrong with renormalization and quantum mechanics as a whole.

[DKFX]

I've brought up in conversation this topic and this specific paper online around a year ago and was getting a lot of heat and people attacking me, but very little substance in their arguments.
Glad to see you giving it more exposure.

[magtovi]

During grad school, my advisor ALWAYS pointed out all this mess.
The theory he was (is) developing for which I contributed a little bit was founded in differential geometry and looked quite promising. I hope it leads somewhere.

[AmaterasuSolar]

I found this fascinating! I will say, My dad was educated by Feynman - and spoke of Him often. I just want to correct the pronunciation of "Feynman..." It is pronounced "Fine man."

[theelectricorigins]

We must learn to differentiate Science (which can only be made about PRESENT here and now) and Mythology which are subjetive interpretation. Scientific methos is about Experiments and Observation.
e.g. Double-Slit Experiment and interference patterns are Science. The Copenaghen or the Many Worlds Interpretation are Mythology!

[tomctutor]

What we are doing here is asserting the current QED is a theoretical description of Lepton-Lepton or Lepton-Hadron interaction.
It could be if someone could construct a proper Quantum Field Theoretical (QFT) model. The closest we get is Feynman, which by the way is used for all sorts of particle interactions quite successfully so has some merit.
The aforementioned QED formula is what in physics we call an *"EMPIRICAL"* formula. As such it involves matching free parameters (the α's coefficients) in a polynomial to experimental measurements. There is nothing "rotten" or "suspicious" about this and Feynman is right and admits its not 100% correct. No one is claiming it is theoretically sound.
To illustrate consider Kepler's EMPIRICAL laws of planetary motion, ;which ere very good at determining the orbits of the planets and moons etc. It took the theoretical genius of Issac Newton to construct a single Universal Law, based on Kepler's work, and with a sound mathematical framework, which was later improved on by Einstein.
The ultimate answer to all this is an extension to the Standard Model from QFT, who knows maybe String Theory or Supersymmetry will give us this. It is fundamentally wrong to rubbish a beautiful and successful model just because someone made an arithmetical error.
I finish by restating Michio Kaku's observation: "If you think my model is wrong, you come up with a better one"!

[LRRPF]

It is interesting that Julian Schwinger after a good deal of chicanery, ignoring Dyson's critique, was effectively dismissed from Harvard, as his genuine creativity came into question. Yet many purport to calculate the anomalous electron magnetic moment to great accuracy, while ignoring the atrocious instability of the Standard Maxwell-only electron. I don't care how accurate they claim the QED anomalous magnetic moment is, it doesn't prove that QED is on sound footing.

[robhannum]

Yeah.. Math is supposed to be the language of physics/science. But just like any language, you can use it to write fiction. Thanks for posting.